[Emacspeak] Re: Introduction & Voice Configuration Questions

[D.J.J. Ring, Jr. <n1ea@xxxxxxxxxxx>]

Oralux.org sells the old IBM Text to Speech modules (EUR 4.19) which work in emacspeak, and https://oralux.org/ has an installer which works in many Linux distros. Additionally, Oralux now sells the Vocalizer Embedded voices in an amazing number of languages and dialects (400) which work in text console and in Orca with satisfying results.  Vocalizer Embedded is selling for EUR 22.10 to 29.81 depending on the quality of the voice from "Compact" to "High".

I use Zoe ML- High because it was easiest for me to hear.

I have it working in console with Debian and Slint / Slackware.

It would be wonderful if it were free to use and modify, but it is what it is and it's very good.

I also wish a speech engine for it would be developed because I find the voices very easy to understand even at fast rates.

I have no connection with oralux.org other than being a happy customer of both the IBM TTS voice which I use in emacspeak and the Embedded voices.

I think it was easiest for me to use the Embedded voices in Orca in GUI and in the CLI but espeak-ng in emacspeak, because I could not figure out how to run regular IBM TTS in emacspeak while running Voxin Embedded in console.  I have a desktop keyboard and I just press num pad 0 and Enter and the console voice stops, type "emacs" and I have emacspeak working. Voxin provides a script called "voxin-say" to hear and test it's modules. voxin-say "text to say" | aplay being the command. voxin-say --L lists the installed voices which must be inserted into and selected in configuration files voxin.ini.

Best wishes,

David

On Fri, Jan 21, 2022 at 12:10 AM Tim Makarios via Emacspeak <emacspeak@xxxxxxxxxxx> wrote:

Hi,

I don't think I'd heard of Larynx before. Have you considered RHVoice as a possibility, if you're going to be writing a speech server for Emacspeak?
https://github.com/RHVoice/RHVoice
I think it lets you adjust pitch and rate, which Emacspeak can make use of if you put the right pipes in the right places (or so I understand).

The licensing of the voices is a little confusing, which is presumably why it ended up in Debian non-free, but I *think* some of the voices (like Alan, as far as I can see) are covered only by the (L)GPL, which is applied to the software as a whole.

As for how to do it, I'm certainly no expert in that, but here's some relevant documentation:
http://tvraman.github.io/emacspeak/manual/TTS-Servers.html

All the best,

Tim
<><

On 21 January 2022 8:24:50 am NZDT, Hendur Saga via Emacspeak <emacspeak@xxxxxxxxxxx> wrote:

Hello Emacspeakers!

This is my first post here, so I thought I might as well introduce myself first. I've been a programmer for over a decade, a moderate Emacs user for a few years, declared "bankruptcy" a few times, and am still using it to this day. I have no visual difficulties at present, but I've found auditory UIs (and UX design in general) rather fascinating.

Unfortunately, the default espeak voice I'm stuck with so far gets very irritating, very quickly. I almost want to punch a wall, sometimes. I'm sure some of you can sympathize. I've looked through the *-voices.el and, for a Linux user, there doesn't really seem like much choice, sadly. If I were to try and add support for, say, Rhasspy's Larynx project (available at https://github.com/rhasspy/larynx), how difficult would such a task be, and what path or resources could I look at to help achieve that?

In the meantime, how do I at least switch espeak voices, within the same language? I have additional MBROLA voices from Debian I can use from the command-line.

Thanks,

Hendursaga

--
Sent from my Android device with K-9 Mail. Please excuse my brevity.

_______________________________________________
Emacspeak mailing list -- emacspeak@xxxxxxxxxxx
To unsubscribe send an email to emacspeak-leave@xxxxxxxxxxx